The lung is a radiosensitive organ and at least half of all the nuclear accidents reported between 1950 and 2000 include lung injuries. Unlike gastrointestinal, skin and hematological injuries, there is a significant window of time (6-8 weeks) between radiation exposure and symptoms of pulmonary dysfunction. We have evidence that mitigating agents can reduce lethality from pneumonitis in rats when started as late as 35 days post exposure. Therefore, biomarkers that predict lung injury in the first 30 days after a radiological event will serve a very important role in identifying and protecting victims o a mass casualty event from lung injury. We have identified doses of total body and whole thorax irradiation that induce lethal pneumonitis in a reproducible manner from 0-100 percent of rats. Using such models, we have tested over 15 assays of lung function and 26 circulating proteins. Based on our results and those of others, we have chosen three classes of minimally invasive and novel biomarkers as candidates to predict lethal pneumonitis. We will test our biomarkers with low (X-rays) and mixed (neutrons and gamma rays) linear energy transfer (LET) sources of radiation that resemble an A-bomb. In Aim 1, exhaled breath and blood markers will be tested. Increase in the gas nitric oxide (NO) in exhaled breath, has been demonstrated to predict radiation pneumonitis. Blood will be used to identify all novel or existing circulating microRNA that may predict lung injury, using an exhaustive next-generation sequencing (microRNA seq) approach. Aim 2 will focus on another area of interest of the RFA, organ-specific imaging. Using single-photon emission computed tomography (SPECT) we will image lungs to measure the decrease in pulmonary perfusion in vivo (known to occur in a specific pattern after radiation) and/or pulmonary cell death that may predict lethal pneumonitis. In Aim 3 we will confirm biomarkers positive for radiation-specific lung injury from Aims 1 and 2 in a second species of rat. Finally, we will confirm our results by conducting a blinded trial in rats to predict radiatio pneumonitis and measure the accuracy of the selected biomarkers. A combination of established models, a high throughput sequencing technique and efficient study design will allow us to complete our aims in a timely manner. We have established models of radiation injury and a team which includes renowned experts to help us reach our goals.